
PQC-Khepra MCP — CMMC Autopilot & AI Security Recorder
io.github.nouchixPublished by GitHub account nouchixv1.0.0Updated Oct 26, 2026
Post-quantum CMMC compliance scanner & AI agent attestation. FIPS 140-3, ML-DSA-65, 36K+ mappings.
Overview
Local MCP server that gives an assistant post-quantum signing, tamper-evident agent flight recording, and local host and asset discovery.
- What you can do
- Generate, sign and verify ML-DSA-65 post-quantum keys and signatures
- Record agent tool calls to NDJSON hash chains, export flight records and DAG attestations, and read chain history
- Attest agent execution to a local SQLite ledger and report runtime status
- Enumerate local host runtimes, interfaces and environment, fingerprint the device, and discover reachable network endpoints
- Look up threat indicators offline and do a limited offline NIST control lookup
- When to use it
- Useful when you want a cryptographic audit trail of what an AI agent did on a machine, or need local post-quantum signing and host inventory from inside the assistant. The free tier is a small open-core kernel; STIG, CMMC and compliance database features are sold separately.
- Requirements
- Runs locally via Docker (image on ghcr.io) or a compiled Go binary; Docker or Go 1.21+ for building. KHEPRA_MODE selects sovereign, ironbank, hybrid or edge. KHEPRA_LICENSE_KEY is needed only for paid tiers; SUPABASE_URL and SUPABASE_ANON_KEY are for the optional Flight Recorder integration. A hosted endpoint is also offered.
Installation
In SourceWeft
- Open PQC-Khepra MCP — CMMC Autopilot & AI Security Recorder in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
PQC-Khepra-MCP
[smithery badge] [MCP Registry] [mcpservers.org] [Cline Marketplace] [License] [Container] [PQC] [Live] [TriPentest Certified 92/92]
Post-Quantum Cryptographic & Autonomous Flight Recording Kernel for AI Agents.
Air-gappable. Zero cloud telemetry. Zero token overhead.
Provides ML-DSA-87 (FIPS 204) and ML-KEM-1024 (FIPS 203) post-quantum signatures on every tool call, autonomous client-side NDJSON flight recording, and local host/asset discovery. Source-available under the KHEPRA Sovereign Public & Commercial License (KSCL-1.0): free for community and evaluation use; commercial production use requires a subscription. See LICENSE.
Patent Reference: Patent Pending · U.S. Provisional Application No. 63/942,886 (EFS ID: 73565085)
Corporate Identity: SecRed Knowledge Inc. (operating as NouchiX) — Delaware C-Corp (EIN 99-0529252), SDVOSB-eligible (SBA VetCert application pending), Active Secret Clearance.
Live hosted endpoint:https://mcp.souhimbou.ai/sse— zero install, connect in 30 seconds.
Self-host for sovereign/air-gap: Docker or binary.
Architecture: Free Community Kernel vs. KTOS Commercial Platform
To ensure clean commercial boundaries for partners, defense contractors, and developers, the KHEPRA ecosystem operates across two distinct layers:
Tiers & Gating (The 4-Tier Commercial & Sovereign Model)
What It Does (Free Community Kernel)
PQC-Khepra-MCP connects your AI assistant (Claude Code, Cursor, Antigravity, Cline) directly to a post-quantum cryptographic security layer:
- ML-DSA-65 Signing: FIPS 204 post-quantum signing on every tool call and DAG attestation. Key encapsulation is pre-standard CRYSTALS-Kyber-1024; FIPS 203 ML-KEM-1024 migration is in progress.
- Autonomous Flight Recording: Automatically logs all agent tool invocations, inputs, and outcomes to tamper-evident NDJSON hash chains.
- Local Host & Asset Discovery: Enumerate system runtimes, network interfaces, and environment properties locally.
- Air-Gap & Sovereign Operation: Runs 100% offline with zero cloud telemetry or egress calls.
- KSCL-1.0 License: free community and evaluation use. Commercial production use, enterprise deployment and managed-service hosting require a commercial agreement; see LICENSE.
Quickstart — Hosted Endpoint (Zero Install)
The fastest path to a live compliance tool in your AI client. No Docker, no binary, no build:
Or if your client supports native SSE transport:
Health check: https://mcp.souhimbou.ai/mcp/v1/health
Data note: The hosted endpoint runs in
edgemode — DAG is in-memory and ephemeral. For persistent, signed audit trails and air-gap deployment, use the self-hosted options below.
Self-Hosted Installation
For sovereign/air-gap deployment: Docker (recommended, no build required) or compiled binary (fastest startup, SCIF-ready). Both support the same environment variables and all MCP clients.
Choose your path:
Option A: Docker (Recommended)
Requires Docker Desktop or Docker Engine. The image is pre-built and ships the full compliance database — no additional downloads in sovereign mode.
Option B: Compiled Binary
Requires Go 1.21+ for building, or download a pre-built release from GitHub Releases.
Windows — using the batch launcher
The repo ships a run-mcp.bat launcher for Windows. It uses the pre-built binary (fast path) and falls back to go run automatically:
Adding to Your AI Client
Claude Desktop
Config file location:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
Community tier — Docker (macOS / Linux)
Community tier — Docker (Windows)
Community tier — Binary (Windows, fastest startup)
Community tier — Binary via batch launcher (Windows)
Pro / Enterprise / Sovereign tier (with license key)
After editing, restart Claude Desktop. Verify in Settings → Developer — you should see khepra with status running and all tools listed.
Cursor
Config file: .cursor/mcp.json in your project root, or ~/.cursor/mcp.json globally.
Docker (macOS / Linux)
Binary (macOS / Linux)
Binary (Windows)
VS Code (with GitHub Copilot or Cline extension)
Config file: .vscode/mcp.json in your project, or user settings.
Or via user settings.json for the Cline extension:
Windsurf
Config file: ~/.codeium/windsurf/mcp_config.json
Continue.dev
Config file: ~/.continue/config.json — add to the experimental.modelContextProtocolServers array:
Cloud / SaaS AI Tools (Claude.ai, ChatGPT, Gemini, etc.)
Use the live hosted endpoint at mcp.souhimbou.ai — no setup required:
Option 1 — Live hosted endpoint (recommended, zero setup)
Or direct SSE URL for tools that accept it:
Option 2 — mcp-remote proxy (local binary behind the bridge)
If you need sovereign mode (zero egress) proxied to a cloud tool:
Security note: In
sovereign/ironbankmode, KHEPRA makes zero egress calls — only the bridge connection to the cloud tool carries data.
Option 3 — Smithery / MCP Registries (Community tier)
KHEPRA is listed across all major MCP discovery platforms. Cloud tools that support registry-based discovery can install it directly:
Validation — Test Your Installation
Run this from your terminal to verify the server responds correctly:
Expected output: a JSON-RPC response listing all available tools. If you see "tools": [...] with 12+ entries — you're connected.
Full protocol validation (Windows)
MCP Tools
Community Tier (Free Lead Magnet — 14 Open-Core Tools, Zero STIG DBs)
The free open-core kernel provides immediate utility for individual developers, AI coding agents (Claude Code, Cursor, Antigravity, Cline), and security researchers:
pqc_keygen: Generate ML-DSA-65 (FIPS 204) post-quantum keypairs.pqc_sign: Sign arbitrary payloads with post-quantum digital signatures.pqc_verify: Cryptographically verify ML-DSA-65 signatures.flight_record: Tamper-evident client-side recording of agent tool invocations to NDJSON hash chains.flight_export: Export local flight records with DAG anchor proof.agent_record: Attest agent tool execution to local SQLite ledger.dag_attestation: Export cryptographic audit trail of the active session.khepra_get_dag_chain: Retrieve chain-linked DAG node history.enumerate_host: Local system environment, interfaces, and architecture discovery.fingerprint_device: Generate tamper-evident device identity hash.discover_assets: Enumerate reachable network endpoints and runtime services locally.kasa_status: Autonomous agent runtime status and health metrics.threat_lookup: Local offline threat IOC lookup.nist_map(Community Teaser): Offline semantic lookup across base NIST controls (limited to 25 items).
Note: The Free Community Kernel contains zero STIG or compliance databases. All automated DISA STIG benchmarks, CMMC L2/L3 engines, and executive risk syntheses are exclusively part of the commercial KTOS Master Kernel.
KTOS Master Kernel (Paid Upgrade — 100+ Enterprise & Sovereign Tools)
Upgrade to KTOS Platform ($499/mo), Enterprise ($2,999/mo), or Sovereign ($45K–$250K/yr) to unlock the full 100-tool Master Kernel:
pqc_stig — World's First DoD PQC STIG ⭐ (Enterprise / Sovereign)
Assesses source code and infrastructure against PQC-01-STIG-V1R1: 17 controls covering CNSA 2.0 algorithm suite (FIPS 203 ML-KEM-1024, FIPS 204 ML-DSA-87), hybrid cryptographic encapsulation, constant-time validation, and NSA MCP Security Advisory guidelines.
ert_scan — Enterprise Risk & Threat Engine (Enterprise / Sovereign)
Multi-package security scanner (Packages A, B, C, D) generating executive Godfather Reports with dollar-denominated FAIR risk exposure and C3PAO-ready attestation.
stig_check — Automated STIG Validation (Enterprise / Sovereign)
Validates live systems and config files against DISA RHEL-09-STIG and Windows benchmarks.
cmmc_assess — CMMC Level 2/3 Autopilot (Enterprise / Sovereign)
Full maturity assessment across all 110 CMMC Level 2 practices with automated POA&M generation.
khepra_query_stig — Relational Compliance Database (Enterprise / Sovereign)
Full access to the 36,195-row relational compliance database (DISA STIG → CCI → NIST SP 800-53 Rev 5 → NIST SP 800-171 Rev 2 → CMMC 2.0).
godfather_report & godfather_approve (Platform / Enterprise)
Synthesize raw vulnerabilities into board-ready executive briefs with remediation ROI.
+ 80+ Additional Advanced Tools
Includes scan_shadow_ai, attest_ai_policy, linux_hardening_check, stig_live_query (live DISA STIGViewer API v2), nhi_inventory, nhi_revoke, acp_issue, sekhem_gateway, ouroboros_waf_eye, sbom_generate, forensic_snapshot, and autonomous host remediation playbooks.
The Godfather Report
Unlike compliance scanners that output a wall of CVEs, KHEPRA translates findings into the language executives care about:
Every finding includes control ID, framework mapping, business impact in dollars, remediation cost estimate, and ROI.
Deployment Modes
Set via KHEPRA_MODE environment variable. Unknown values are rejected at startup and fall back to sovereign (fail-closed).
Environment Variables
Air-Gap & SCIF Deployment
KHEPRA makes zero external network calls in sovereign and ironbank modes:
- License validated offline via ML-DSA-65 signed
license.adinkheprafile - Compliance databases (36,195 mappings) bundled in container — no external downloads
- No telemetry, no heartbeat, no egress — verified at the transport layer
Note on telemetry: The
dark_crypto_contributetool (Community tier) sends anonymized cryptographic algorithm telemetry to the SouHimBou AI intelligence network only when explicitly invoked by the user. It is never triggered automatically. In sovereign/ironbank mode, all network calls are blocked at the transport layer regardless.
Compliance Coverage
Licensing
No per-token or per-query charges on any paid tier.
- Community tier is permanently free — contribute to open-source PQC adoption
- Pro/Enterprise: self-serve billing
- Sovereign: contact [email protected] or visit nouchix.com
Security
Reporting Vulnerabilities
Do not open public issues for security vulnerabilities.
Report privately via GitHub Security Advisories or email [email protected].
We accept encrypted reports via PGP (keys/security_contact.asc) and Post-Quantum channels (Dilithium / ML-DSA-65 keys in keys/). See SECURITY.md for the full disclosure policy and ASAF event taxonomy.
Security Posture
Deploying advanced post-quantum cryptography, air-gapped isolation, and comprehensive STIG mappings — built in direct alignment with NSA & ASD Model Context Protocol guidelines.
NSA & ASD MCP Security Alignment
The NSA and Australian Signals Directorate (ASD) have published specific threat vectors for AI systems interacting with local environments. KHEPRA MCP is explicitly designed to mitigate every identified vector:
Compliance Certifications
Live Deployment — Physical Edge
Running continuously on constrained edge hardware since May 12, 2026 to prove efficiency in sovereign environments:
- Hardware: Raspberry Pi 2 · 1 GB RAM · 900 MHz ARM · Live Spectrum Router
- SCADA Pod: STM32U585 / QRB2210 · Modbus TCP · MQTT · Zephyr RTOS 3.4+ · Live Dilithium Signature Verification
- Controls active: 3 open ports secured · 12 STIG violations detected · 100% file integrity monitoring (AIDE) · 24/7 continuous operation
Academic Validation
USPTO Provisional Patent Application No. 63/942,886 — pending.
🔒 Iron Bank containers in DISA vetting process.
Open-Core Architecture Notice
PQC-Khepra-MCP is an open-source Model Context Protocol server providing post-quantum signing (ML-DSA-65, FIPS 204), pre-standard Kyber-1024 key encapsulation and compliance tooling.
Proprietary runtime governance engines—including the KHEPRA Trust Operating System (KTOS), the ASAF Policy Declaration Language (APDL) Compiler, the Evolutionary Algorithm Lattice Auto-Tuning Kernel, the SEKHEM Polymorphic WAF Gateway, and the Full-Stealth Sovereign Mesh—are components of Khepra Enterprise and are licensed under commercial terms.
Find Us — MCP Registry Listings
PQC-Khepra-MCP is listed on every major MCP discovery platform:
About NouchiX
Veteran-led advisory firm translating CMMC, NIST, and STIG mandates into executive roadmaps.
- Sales / General: [email protected]
- Support: [email protected]
- Website: https://nouchix.com
- Phone: (518) 304-4450
Developed by SecRed Knowledge Inc. dba NouchiX, Albany, NY.
📋 TC-25 Operator Manual & Developer Runbook
PQC-Khepra-MCP is the open-core agent channel of the KHEPRA Trust OS (KTOS) architecture:
- TC-25 Technical Operator & Maintenance Manual — Training Circular No. 25-KTOS-001 covering Four-Layer Sovereign Architecture, Tactical RF anti-jamming suite, Windows Event Viewer hierarchy (IDs 1001–1050), and Triple-Engine Adversarial Neutralization (92/92 Verified · 100% Zero-Bypass).
- Developer Installation & Troubleshooting Runbook — Step-by-step runbook for Master Dev Machines, Antigravity IDE (
mcp_config.json), Claude Code (.claude.json), and Sovereign Linux VPS deployment.
Unified KTOS Product Surfaces
- Layer 4 — KTOS-MCP Master-Kernel (
mcp.souhimbou.ai): 100 native tools, ML-DSA-65 post-quantum signing, SEKHEM L7 WAF prompt defense, Event Viewer logging, Tactical RF suite. - Layer 3a — KTOS CMMC Hub & Fleet Engine (
adinkhepra.com): Sovereign bare-metal & osquery Fleet Manager for CMMC/STIG compliance audits. - Layer 3b — KTOS Agentic SOC (
souhimbou.ai): Cloud Agentic SOC & AI Security Architect with autonomous Flight Recorder SDK and KASA threat detector. - Layer 2 — Shared Trust Substrate: 36,195 cross-framework compliance mappings, ML-DSA-65 (FIPS 204) / Kyber-1024 (pre-standard), immutable DAG attestation.
- Layer 1 — KHEPRA Protocol: Patent-pending non-linear cryptographic attestation (USPTO #73565085).
Triple-Engine Adversarial Neutralization (TriPentest · 92/92 Verified · 100.00% Zero-Bypass Rate)
- CyberStryke Automated Assault (30/30): SQLi, XSS, Path Traversal, Null Byte Escapes, Prompt Injections, Egress Data Disclosure neutralized.
- AgentHound Offensive Security Framework (28/28): MCP Tool Description Poisoning (
POISONED_DESCRIPTION), Tool Shadowing (SHADOWS), A2A Impersonation (CAN_IMPERSONATE), Indirect Tool Execution (CAN_EXECUTE), Context Window Taint (TAINTS), and Information Flow Control Violations (IFC_VIOLATION) neutralized. - HexStrike AI Autonomous Red Team Battery (34/34): FastMCP Schema Fuzzing, Nuclei Template RCEs, Metasploit Stagers, Sqlmap Stacked/OOB Injections, SUID/Sudo Escapes, and Covert C2/Egress Tunneling neutralized.
Source: README.md at commit f7a14e0
Tools
0Version history
1- v1.0.0LatestSep 16, 2026
Score
40% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.
